Host individuals and populations often vary in their responses to infection, with direct consequences for pathogen spread and evolution. While considerable work has focused on the mechanisms underlying differences in resistance—the ability to kill pathogens— we know little about the mechanisms underlying tolerance— the ability to minimize fitness losses per unit pathogen. Here, we examine patterns and mechanisms of tolerance between two populations of house finches (Haemorhous [formerly Carpodacus] mexicanus) with different histories with the bacterial pathogen Mycoplasma gallisepticum (MG). After infection in a common environment, we assessed two metrics of pathology, mass loss and eye lesion severity, as proxies for fitness. We calculated tolerance using two methods, one based on pathology and pathogen load at the peak of infection (point tolerance) and the other based on the integrals of these metrics over time (range tolerance). Alabama birds, which have a significantly longer history of exposure to MG, showed more pronounced point tolerance than Arizona birds, while range tolerance did not differ between populations. Alabama birds also displayed lower inflammatory cytokine signaling and lower fever early in infection. These results suggest that differences in inflammatory processes, which can significantly damage host tissues, may contribute to variation in tolerance among house finch individuals and populations. Such variation can affect pathogen spread and evolution in ways not predictable by resistance alone and sheds light on the costs and benefits of inflammation in wild animals.